Better Process for Alteryx Server backup and Restore

[joeGade]

A better process Needed for Alteryx Backup and restore.

1. Why do we have to stop the service to do a backup,? We cannot stop a service in Enterprise platform just to do a backup.  we are not able to take backups because our jobs will be running 24X7 We Run jobs on Alteryx Server,
2. All of a Sudden our Windows server crashed and we lost all the data, The backup we have of Mongo DB was not perfectly restored on the new server which breaks the workflow, connections etc  due to encryption....
3. what is the purpose of backup if we cannot restore on a new server in case of DR

Please fix it ASAP,

[sgabriel62]

@Gade

Look at it from the backup perspective -  taking a backup of a running database is what we call "fuzzy backups".  Fuzzy backups are no guarantee for success, never has never will.  So we have to do things the long way, we do need to put the database into a static state.  As it currently stands today, Mongodb does not support the opportunity to perform "hot backups" or the ability to quiesce the DB so that a snapshot can even be performed.  I've raised this same question as well and there is only one answer.   MongoDB Dumps require the service to be stopped.  Same for MongoDB Restores.

Painful as it may seem, It is what it is!!!

If your MongoDB is under 100GB in size, you are looking at a very quick dump process, and this process can be schedule in between a quiet time zone in your workflow schedules.   If you database is > 100GB you are looking at an hour exactly for the dump to complete.

My environment, similar to yours is a Global real-time environment, but my team and I are working on reducing the database to a manageable size so that the Dumps can be executes in minutes vs. hours.   From here we slide in the routine in between the workflow schedules.

The business needs to become aware of the need.  We are gradually announcing our intentions by providing a stable environment and an increased number of scheduled maintenance activities where a backup is a Pre-req.  to include System patching at the OS level.   Your down right - why not back it up.

To make a long story short - Successful DR is your ability to backup & restore.  Successful backups is dependent on Database size and manageability.  The more backups you run - the more of a baseline you establish and sooner than later you have created a maintenance window.

Good Luck - let me know where I can help you in any way

Steve

[Mallik_788]

Why not get look towards direction of support of Other DB's that can handled Hot or Fuzzy Back-ups effective,y

[DD266]

Hi,

Even in 2020 do we need to stop the service for Backup? or is there any new update?

Also is it possible to take an incremental backups after a full back to save storage.

Thank you!